Poolside

Tin tức, mô hình và bản phát hành AI mới nhất từ Poolside. ['Laguna S 2.1', 'Laguna XS', 'Laguna XS.2']

Nghiên cứu 🇺🇸

Latest innovations in LLM architectures: shared KV caches, per-layer embeddings, and compressed attention

New open LLMs increasingly focus on efficient long-context processing. Gemma 4 (Google) uses KV tensor sharing between layers to reduce cache size, as well as per-layer embeddings to improve parametric efficiency. Laguna XS.2 (Poolside) employs per-layer attention budgeting, and DeepSeek V4 introduces mHC and compressed attention mechanisms.

Google/DeepMind

Poolside

DeepSeek

Technology Innovation Institute

Sebastian Raschka27.07 · 13:06

Mô hình 🇺🇸

Poolside Releases Laguna S 2.1 — an Open-Weight Model for Agentic Coding That Punches Above Its Weight Class in SWE-Bench Multilingual

Poolside released Laguna S 2.1, an open-weight model with 11.8 billion parameters and 8 billion active per token, designed for agentic coding. Based on a Mixture-of-Experts (MoE) architecture, it supports up to 1 million tokens of context and can run on a single NVIDIA DGX Spark. Laguna S 2.1 ranks first among open models with known size on Terminal-Bench 2.1 (70.2%) and SWE-Bench Multilingual (78.5%), outperforming many larger competitors.

Poolside

DeepSeek

NVIDIA

Tencent

Moonshot AI

Alibaba/Qwen

Anthropic

MarkTechPost24.07 · 05:04

Xem thêm ↓